Daily Poll Watch: Romney likely will carry more states than Obama but still lose the race

Please note my use of the word “likely” in the headline above. That means I’m not flat-out predicting anything. I’m just assessing the situation as it looks now, a week before Election Day.

At this writing, Mitt Romney seems to be running at least fairly well in 23 states, while President Obama is doing at least fairly well in 18 states. And the other nine states are said to be still up for grabs.

I think it’s reasonable to guess that Romney will carry at least three of those so-called swing states, which would give him a majority of the 50 states.  But I doubt that he’ll prevail in the swing states he most needs to put him over the top in the Electoral College.

It might also turn out that Romney will win the national popular vote and still fail to garner enough electoral votes to capture the presidency. But I think there’s only one chance in three or four of that happening.

The bottom line here is that the polling analysts I most respect, Sam Wang and Nate Silver, still see a strong likelihood that Obama will get more than the 270 electoral votes he needs to win. I have no reason to question their projections, nor do I know of any other serious analysts who have come to different conclusions using similar methods of analysis.

Wang’s current estimate is that Obama will win 303 electoral votes. Silver puts the figure at 295 electoral votes.

If there’s any shift of electoral strength toward Romney over the next five or six days, which is not impossible by any means, I’m sure that Wang and Silver will catch it and revise their projections accordingly. Those projections may not be in complete agreement with each other, but it’s unlikely that one will see Obama winning while the other one doesn’t.

Silver’s conservative critics seem not to understand that he’s merely reporting on the numbers in the various state polls. He examines virtually all the polls in any given state, even the polls from Rasmussen, arrives at an average, assesses the likelihood of a significant shift, and then assigns the appropriate number of electoral votes.

As he put it in an interview yesterday with Politico: [T]his is not wizardry or rocket science. All you have to do is take an average, and count to 270. It’s a pretty simple set of facts.”

Silver also notably said this: “Romney, clearly, could still win.” But, of course, he’s not going to say that a Romney victory is likely unless and until it is likely.

We’ll probably know by Saturday or Sunday if there’s any significant shift afoot. If there is, then we’ll likely declare on Monday that all bets are off.



  1. Pat Cunningham likely will eat his words and feel very silly on November 7 after Obama is defeated.

    Please note my use of the word “likely” in the sentence above. That means I’m not flat-out predicting anything. I’m just assessing the situation as it looks now, a week before Election Day.

    The liberals out here on the West Coast are totally freaking out. They see Obama’s loss as more and more certain. This morning my dentist asked me, almost tearfully, if I thought there would be riots in the streets when Romney wins. (Yes, there may well be, in the biggest cities of the blue states.)

    I do predict that Obama’s defeat will cause a tidal wave of finger-pointing in the Democratic party, and the party will eventually split into two when the progressives secede and form their own loony party that sails right over a cliff in the 2014 elections.

    Just as the Emperor Palpatine from Star Wars said to Darth Vader: “Everything is proceeding as I have foreseen.”

  2. I haven’t analyzed Wang’s projections, but I have looked into Silver’s, and he has two fundamental problems with his methodology.

    1. His weighting of certain polls over others – From a purely scientific viewpoint, his weights don’t appear to add up. I tend to agree with Nate that Rasmussen leans too far to the right, and they deserve less weight comparatively to other pollsters. However, he appears to consistently give Democratic-leaning PPP polls the most weight, even more than those he generally trusts such as SurveyUSA, Pew, and Gallop.

    2. The polls themselves are somewhat skeptical. Predominately, its the Democratic sample size that is often too close to the 2008 results when at the same time there appears to be universal consent that Republican turnout and motivation will be much higher than what they had in 2008 and comparatively Democratic apathy is also more predominate. So while a sample size of D+3 appears about right, I’ve seen several polls that show a tie, but the sample is D+9 which is completely unrealistic, meaning that in such a poll the Republican actually has the lead.

    So what Nate has is a fundamentally flawed data set, and yet he builds his projections around the model. So Pat rightfully asks “Where is the Republican equivalent to Nate Silver?” The answer is I don’t know. There is one attempt out there called “unskewed polls” but they move so far to the right where its hard to take them serious.

    The general Republican talking point has been around my point #2 above, yet there doesn’t appear to a be person who publicly tallys this information from a worthy right wing perspective. Heck, Nate the Great only came into prominence 4 years ago. Perhaps by 2016 there will be an established right wing equiv. (I’m sure the RNC and DNC have their own internal polling experts, but they obviously don’t release their data and strategies to the general public.)

  3. Despicable, dishonest and desperate Mitt still trying to win Ohio by touting Chinese Jeep manufacturing.

Leave a Reply

Your email address will not be published. Required fields are marked *